NeuroToxicology 26 (2005) 651–659

Role of (PON1) Status in Pesticide Sensitivity: Genetic and Temporal Determinants Clement E. Furlong1,*, Toby B. Cole1,2, Gail P. Jarvik1, Christina Pettan-Brewer1, Gary K. Geiss1, Rebecca J. Richter1, Diana M. Shih3, Aaron D. Tward3, Aldons J. Lusis3, Lucio G. Costa2 1Genome Sciences and Medicine, Division of Medical Genetics, University of Washington, Box 357720, Seattle, WA 98195-7720, USA 2Environmental and Occupational Health Sciences, University of Washington, Seattle, WA 98195-7720, USA 3UCLA, Microbiology and Molecular Genetics, Los Angeles, CA, USA

Received 14 June 2004; accepted 5 August 2004 Available online 29 September 2004

Abstract

Individual differences in detoxication capacities for specific organophosphorous (OP) compounds are due largely to differences in catalytic efficiency or abundance of the HDL-associated , paraoxonase (PON1). First, we provide evidence that children less than 2 years of age represent a particularly susceptible population for OP exposure due to low abundance of PON1 and variable onset of plasma PON1 activity. Second, we describe studies examining the neurotoxic effects of chronic, low-level OP pesticide exposure in mice. PON1 knockout (PON1/) and wild-type mice were exposed chronically (PN4 to PN21) to low levels of chlorpyrifos oxon (CPO). Endpoints included activity, histopathology, expression, and behavior. Even at PN4, when PON1 levels were low in wild-type mice, PON1/ mice were more sensitive to inhibition of brain cholinesterase by CPO. At PN22, and persisting as long as 4 months, chronic developmental exposure to 0.18 mg/kg/d or 0.25 mg/kg/d CPO resulted in perinuclear vacuolization of cells in a discrete area of the neocortex and irregular distribution of neurons in the cortical plate, with an increase in the number of affected cells at 0.25 mg/kg/d. Third, we describe a transgenic mouse model in which human transgenes encoding either hPON1Q192 or hPON1R192 were expressed at equal levels in place of mouse PON1. The developmental onset of expression followed the mouse time course and was identical for the two transgenes, allowing these mice to be used to assess the importance of the Q192R polymorphism during development. Adult mice expressing hPON1R192 were significantly more resistant than hPON1Q192 mice to CPO toxicity. Our studies indicate that children less than 2 years old, especially those homozygous for PON1Q192, would be predicted to be particularly susceptible to CPO toxicity. # 2004 Elsevier Inc. All rights reserved.

Keywords: Paraoxonase; PON1; Organophosphorous compound; Polymorphism

INTRODUCTION with a focus on the high-density lipoprotein (HDL)- associated enzyme paraoxonase (PON1). PON1 exhi- This review focuses on specific genetic and temporal bits considerable variation among individuals due to a factors that influence sensitivity to the commonly used coding region polymorphism that alters its activity and insecticides chlorpyrifos (Dursban1) and diazinon, to differences among individuals in plasma enzyme levels, which also vary temporally during develop- * Corresponding author. Tel.: +1 206 543 1193; ment. Genetic and temporal variability in this enzyme fax: +1 206 543 3050. affects not only sensitivity to pesticides, but also risk E-mail address: [email protected] (C.E. Furlong). for vascular disease and pharmacokinetics of drug

0161-813X/$ – see front matter # 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.neuro.2004.08.002 652 C.E. Furlong et al. / NeuroToxicology 26 (2005) 651–659 metabolism, due to its ability to hydrolyze substrates within a genetic class, e.g., within the low metabolizers, other than the organophosphorous (OP) insecticides. there was wide variability in enzyme activity levels in plasma. A refinement of this two-substrate approach for assessment of PON1 activity, that uses diazoxon instead EARLY STUDIES of phenylacetate (Richter and Furlong, 1999), is dis- cussed further. Mazur (1946) first described enzyme activity in plasma that was capable of hydrolyzing OP compounds. Careful and insightful studies by Aldridge (1953a, 1953b) described an that hydrolyzed certain MOLECULAR BASIS OF THE PON1 aromatic esters and OP compounds like paraoxon, the POLYMORPHISM active metabolite of . Aldridge designated this esterase as A-esterase. In addition, Aldridge described a The molecular basis of the PON1 activity poly- B-esterasethat,incontrasttoA-esterase,wasinhibitedby morphism was not elucidated until the human PON1 . His hypothesis that the same cDNAwas isolated and sequenced (Adkins et al., 1993; enzyme, A-esterase, hydrolyzed both phenylacetate Hassett et al., 1991). Two interesting pieces of infor- and paraoxon was proven conclusively only recently, mation came out of the initial sequencing effort. First, with the demonstration that recombinant paraoxonase/ it became evident that PON1 retained its signal (PON1, EC 3.1.8.1) catalyzed both activities sequence, having only the initiator methionine residue (Sorenson et al., 1995). The plasma enzyme paraoxonase removed during secretion. Second, two research groups (PON1) is the A-esterase described by Aldridge, whereas reported independently that variability in the amino B-esterase activity is represented by the plasma choli- acid present at position 192 [glutamine (Q) or arginine nesterases. A survey of the plasma levels of A-esterase (R)], due to a single nucleotide polymorphism, was activityindifferent speciesshowed widevariation among responsible for the PON1 activity polymorphism the species surveyed (Brealey et al., 1980; Costa et al., (Adkins et al., 1993; Humbert et al., 1993). The con- 1987). Species exhibiting higher levels of plasma A- sequences of the PON1Q192R polymorphism for OP esterase activity were also more resistant to the toxicity detoxication are discussed in more detail further. of OP compounds than were species with lower enzyme Another polymorphism was identified that resulted levels. This observation suggested that inter-species dif- in an amino acid substitution [leucine (L)/methionine ferencesinsensitivitytoOPcompoundsareduelargelyto (M)] at codon 55. Some reports have noted an associa- variability in their plasma levels of A-esterase. In tion between the PON1M55 alloform and low PON1 humans, plasma paraoxonase levels vary by at least activity (Blatter Garin et al., 1997; Brophy et al., 13-fold (Davies et al., 1996). In human populations, early 2001b; Mackness et al., 1998). It should be noted that studies examining paraoxonase levels revealed a poly- initially there was a discrepancy in the literature in the morphic distribution of paraoxonase activity, with either nomenclature of these polymorphisms. Whereas some bi- or trimodal distributions reported, depending on the published reports numbered the amino acids from the particular assay used (Eckerson et al., 1983; Geldmacher first residue of the mature protein (Adkins et al., 1993), von Mallinckrodt and Diepgen, 1988; Mueller et al., others numbered the amino acids from the initiator 1983). While the enzyme responsible for the activity methionine (Hassett et al., 1991; Humbert et al., 1993). had not yet been identified, differences in gene frequen- Thus, early papers that report the position of the cies among different ethnic groups were easily seen with polymorphisms as 191 Q/R and 54 L/M are referring even a simple assay. Over the years, the assays used to to the same polymorphisms reported by others to be at investigate the paraoxonase activity polymorphism were positions 192 Q/R and 55 L/M. This issue was finally improved (Ortigoza-Ferado et al., 1984). A major settled during a conference on OP in advance was made by Eckerson et al. (1983),who Dubrovnik, Croatia, where it was decided to number developed a two-substrate assay where the rate of para- the amino acids beginning with the initiator methionine oxon hydrolysis for each plasma sample was measured (i.e., 192 Q/R and 55 L/M) (La Du et al., 1999). The and plotted against its corresponding rate of phenylace- nomenclature for the many new polymorphisms tate hydrolysis. This approach clearly separated the low revealed recently (Jarvik et al., 2003) needs also to metabolizers from the high metabolizers, and came close be standardized. For example, the SP1 to resolving the presumed heterozygotes from the high polymorphism is reported alternatively to be at nucleo- metabolizers. It became clear from this study that even tide position 107 C/T or 108 C/T. C.E. Furlong et al. / NeuroToxicology 26 (2005) 651–659 653

OTHER SUBSTRATES AFFECTED BY THE Fig. 1). This analysis, like the earlier analysis reported PON1Q192R POLYMORPHISM by Eckerson et al. (1983), clearly pointed out the large inter-individual variability in plasma PON1 levels. Davies et al. (1996) extended the two-substrate Fundamental biochemical principals dictate that the approach to a number of other OP compounds, includ- rate of metabolism of a given toxin should be propor- ing chlorpyrifos oxon, diazoxon, soman and sarin. tional to the concentration of the enzyme involved in its Interestingly, the effect of the activity polymorphism detoxication. Thus, it should have been clear early on was reversed for soman and sarin. While PON1R192 that plasma levels of PON1 would be important for hydrolyzed paraoxon at a faster rate than did epidemiological studies examining the relationship of PON1Q192, the opposite was true for soman and sarin, PON1 genetics to risk of disease or exposure. Despite under the conditions used for the in vitro assays. Davies early knowledge that both PON1Q192R genotype and et al. (1996) also reported that PON1R192 hydrolyzed PON1 levels are important in governing rates of toxin diazoxon less well than did PON1Q192. As described metabolism, many epidemiological studies have been further, this observation was the result of the supra- carried out where only single nucleotide polymorph- physiological salt concentration (2 M) used in the in isms have been examined (reviewed in Brophy et al., vitro assays. At physiological salt concentrations, the 2002). Since very few of these studies included a two alloforms had equivalent catalytic efficiencies of determination of PON1 levels (ignoring the significant hydrolysis of diazoxon (Li et al., 2000). contribution that differences in plasma enzyme levels Recently, rates of hydrolysis of a number of drugs play in determining risk), they are of limited value in that contain lactone-rings or cyclic carbonates have estimating disease association with PON1. For epide- been examined (Biggadike et al., 2000; Billecke et al., miological studies where the PON1Q192R polymorph- 2000). The PON1Q192R polymorphism was shown to ism is suspected of affecting the metabolism of the influence the rates of some of the drugs/prodrugs compound in question (e.g., oxidized lipids, OP com- examined. Activation of the antibacterial prodrug, pounds, or drugs), PON1 levels should be measured in prulifloxacin to the active quinolone antibiotic addition to ascertaining genotype at PON1Q192R. The NM394 by PON1 occurred at a higher rate with the combination of determining genotype at PON1Q192R PON1R192 alloform, which also hydrolyzed thiolac- and measuring PON1 levels in plasma has been desig- tones more readily (Billecke et al., 2000; Tougou et al., nated as determining an individual’s ‘‘PON1 status’’ 1998). Inactivation of glucocorticoid g-lactones by (Li et al., 1993; Richter and Furlong, 1999). PON1 confines their mode of action to the sites of application, preventing serious systemic side effects (Biggadike et al., 2000). PON3 appears to be respon- sible for the hydrolysis of statin lactones and the diuretic (Draganov and La Du, 2004). In contrast, the PON1Q192 alloform was more effi- cient at reducing lipid-peroxides in homogenates of human atherosclerotic lesions (Aviram et al., 2000). A considerable amount of research still needs to be carried out to understand the physiological conse- quences of both variation in PON1 levels and the position 192 Q/R polymorphism with respect to all Fig. 1. Determination of PON1 status. Plot of diazoxonase vs. paraoxonase of the many different activities catalyzed by PON1. activities in plasma of carotid artery disease cases and controls, coded for PON1Q192R genotype (determined by PCR) (Jarvik et al., 2003; Richter and Furlong, 1999) Note that the two-substrate assay provides an accurate

inference of PON1Q192R genotype as well as the level of plasma PON1 PON1 STATUS AND EPIDEMIOLOGICAL activities (PON1 status). Because it is a functional analysis, it provides a STUDIES 100% accurate determination of the functional genomics of PON1 status. Newly discovered SNPs explain why some individuals have lower PON1

activity than would be predicted by PON1Q192R genotype alone. Individuals Two-substrate analysis using the substrate pair para- 1–4 genotyped as heterozygotes (Q/R192), however, their enzyme analysis oxon and diazoxon turned out to provide a complete indicated homozygosity for Q or R. Complete sequencing of their PON1 resolution of the three PON1 phenotypes and provided revealed mutations in one allele, resulting in only one alloform of PON1 in their serum. (Note the large variability in PON1 levels, even among a very accurate means of determining PON1Q192R individuals of the same Q192R genotype) (figure reproduced from Jarvik et functional genotype (Richter and Furlong, 1999; al., 2003). 654 C.E. Furlong et al. / NeuroToxicology 26 (2005) 651–659

A recent study examined how well PCR genotyping oxon or chlorpyrifos. Protection was also observed corresponded with PON1 status, as measured using the when purified rabbit PON1 was injected post-exposure two-substrate functional assay (Jarvik et al., 2003). A or 24 h prior to exposure, indicating that administration number of individuals in the study had discrepancies of purified or recombinant PON1 would be useful for between their PON1Q192R genotype and their pheno- ameliorating or even preventing adverse consequences typic assignment in the functional assay (Fig. 1). Only of exposure to OP compounds. after sequencing the entire PON1 genes from these Whereas higher PON1 levels were demonstrated individuals was it discovered that a number of them had clearly to be protective, determining whether low only one functional PON1 allele, due to splicing or levels of PON1 would result in greater sensitivity mutation including premature truncation defects in the was not possible until the development of PON1 other allele. Since PON1Q192R genotyping could not knockout mice, generated by Drs. Jake Lusis, Diana pick up these defects, it provided an inaccurate esti- Shih and co-workers (Shih et al., 1998). Knocking out mate of PON1 status in these individuals, and would the mouse PON1 gene resulted in a dramatic increase give an unreliable estimate of that individual’s risk for in sensitivity to chlorpyrifos oxon exposure and a sensitivity to OP compounds or for vascular disease. modest increase in sensitivity to chlorpyrifos exposure, These examples provide yet another reason for carry- as assessed by measuring brain cholinesterase inhibi- ing out functional analyses of PON1 activity using tion. Dermal exposures to levels of chlorpyrifos oxon high-throughput enzyme assays. All of the 192 known that produced no symptoms of cholinergic effects and single nucleotide polymorphisms in PON1 could be minimal inhibition of brain cholinesterase in wild-type identified in as individual and still not provide the mice were unexpectedly lethal to the PON1 null mice. information gleaned from the two-substrate determi- Similar results were observed when the knockout mice nation of PON1 status. were exposed to diazoxon (Li et al., 2000). Dermal exposure to 2 or 4 mg/kg diazoxon produced no mea- surable effect in wild-type mice, but was lethal to the ROLE OF PON1 IN OP DETOXICATION PON1 knockout mice, and exposure to 1 mg/kg dia- zoxon had significant adverse effects in the knockout While it was assumed that high levels of PON1 mice without measurably affecting the wild-type mice. would protect against exposure to specific OP com- Hemizygous mice, with only one PON1 allele, exhib- pounds, only a single experiment that directly ited intermediate sensitivity. Exposure of the PON1 addressed this question had been reported prior to knockout mice to paraoxon, however, produced an 1990. Main (1956) reported that injection of partially unexpected and initially puzzling result. They were purified PON1 into rats increased their resistance to not any more sensitive than wild-type mice to paraoxon paraoxon. This observation was confirmed and exposure. extended through a series of experiments begun in Further experiments demonstrated that resistance of our laboratory in 1990. Injection of purified rabbit the PON1 knockout mice to diazoxon was restored by paraoxonase into rats increased their resistance to injection of purified PON1R192 or PONQ192 alloforms, paraoxon exposure (Costa et al., 1990). Following this with either alloform providing equivalent protection second demonstration that injected PON1 could (Li et al., 2000). Resistance to chlorpyrifos oxon was increase resistance to paraoxon, we switched to a also restored; however, the PON1R192 alloform pro- mouse model for two major reasons. First, many more vided significantly better protection that did the experiments could be carried out with a given quantity PON1Q192 alloform. Neither alloform provided protec- of purified PON1. Second, it was clear that genetics tion against paraoxon exposure. were much further advanced in mice than rats, with the capability existing for engineering mice with altered PON1 levels. CATALYTIC EFFICIENCY IS THE KEY Injection of purified rabbit PON1 into mice 4 h prior TO PROTECTION to exposure dramatically increased their resistance to chlorpyrifos oxon (Li et al., 1993). An increase in All of these observations were explained by an resistance to the parent compound, chlorpyrifos, was examination of the catalytic efficiencies of the also observed (Li et al., 1995). These experiments PON1Q192 and PON1R192 alloforms for the hydrolysis demonstrated clearly that high levels of plasma para- of each of the OP compounds (Li et al., 2000). Both oxonase could protect against exposure to chlorpyrifos alloforms hydrolyzed diazoxon with equivalent cata- C.E. Furlong et al. / NeuroToxicology 26 (2005) 651–659 655 lytic efficiencies (Vmax/Km = 77). Even though INTERPRETING THE CONSEQUENCES OF PON1R192 hydrolyzed paraoxon with a nine-fold PON1 GENETIC VARIABILITY greater efficiency (Vmax/Km = 6.27) than did PON1Q192 (Vmax/Km = 0.71), the catalytic efficiency was not While there was some protection afforded by PON1 sufficient to provide protection against exposure to against the respective parent compounds, the protective paraoxon in vivo. The better protection provided by effects of PON1 were most striking with the oxon PON1R192 against chlorpyrifos oxon exposure in vivo forms of chlorpyrifos and diazinon. The parent OP was explained by its catalytic efficiency (Vmax/Km = compound is converted to its more toxic oxon form in 256) being higher than that of PON1Q192 (Vmax/Km = the liver, by cytochrome P450-mediated oxidative 150). Pond et al. (1995) had predicted that PON1 desulfuration, and the oxon form serves as the direct would protect against exposure to chlorpyrifos oxon, substrate for PON1. However, direct exposure to the but would not protect against exposure to paraoxon, oxon forms of the OPs in the field is also possible. This based on in vivo rates of hydrolysis of these two caused us to revisit a survey that we had published compounds. some years ago that summarized the percentages of More recent studies with mice expressing oxon found in foliar residues (Yuknavage et al., 1997), either human PON1Q192R alloform (on a PON1 which varied from a few percent to as much as 97% of knockout background) are consistent with the the OP residue (Table 1). While values were not results described earlier. Mice expressing human reported for chlorpyrifos, a California EPA survey of PON1Q192 were nearly as sensitive as the PON1 the oxon content of drift residues during spraying of knockout mice to chlorpyrifos oxon exposures, while chlorpyrifos in Tulare County noted significant oxon mice expressing human PON1R192 exhibited signifi- content in most of the samples analyzed (California Air cantly greater resistance to these exposures (unpub- Resources Board, 1998). Since chlorpyrifos oxon inhi- lished results). This observation is significant in that bits at least 1000 times more PON1Q192 homozygotes comprise as much as 50% of rapidly than chlorpyrifos (Huff et al., 1994), even a populations of Northern European origin (Brophy et small percentage of oxon content is important with al., 2002). respect to an individual’s PON1 status. These observations provide important guidelines for engineering recombinant PON1 for therapeutic appli- cations. For PON1 to protect against toxicity of a given REGULATION OF PON1 EXPRESSION OP compound (or other PON1 substrate), the catalytic efficiency in vivo must be sufficiently high, or no The wide variability in plasma PON1 levels among protection will be provided by PON1, regardless of individuals is due at least in part to polymorphisms that whether the source of PON1 is endogenous or injected have been identified in the 50 regulatory region of (purified native or recombinant). PON1. Our research group and two others have exam-

Table 1 Oxon levels in total pesticide residues taken from dislodgeable leaf foliar residue and dermal exposure studies Study Pesticide (units) Oxona Thioate Total OP Oxon (%) Ralls et al. (1966) Diazinon (ppm)b 0.05 0.25 0.3 17 Kansouh and Hopkins (1968) Diazinonc NDd ––ND Wolfe et al. (1975) Parathion (ng/cm2)b 8 106 114 7 Nigg et al. (1977) Ethion (ng/cm2)b 42 285 327 13 Spear et al. (1977a) Parathion (ng/cm2)b 84 29 113 74 Parathion (mg)e 145 39 184 79 Spear et al. (1977b) Parathion (ng/cm2)b 229 8 237 97 Costello et al. (1989) Malathion (mg)e 659 2301 2960 22 Table modified from Table 1 of Yuknavage et al. (1997). a Based on the highest value reported in study. b Foliar residue measurement. c Units or values not given in study. d ND: none detected. e Dermal monitoring measurement. 656 C.E. Furlong et al. / NeuroToxicology 26 (2005) 651–659 ined the effects of some of these polymorphisms on mining sensitivity to chlorpyrifos/chlorpyrifos oxon plasma PON1 levels (Brophy et al., 2001a, 2001b, and diazinon/diazoxon exposures and prompted us to 2002; Leviev and James, 2000; Suehiro et al., 2000). re-examine one of our earlier observations of low All three groups concluded that the C-108T poly- PON1 levels in newborns (Mueller et al., 1983). It morphism had the greatest affect on plasma PON1 became of interest to define the time course of appear- levels, with the C-108 allele producing about twice ance of PON1 activity in the plasma of newborns. We the level of plasma PON1 as the T-108 allele. The C- found that individual infants approached plateau levels 108T polymorphism is in the middle of a consensus of PON1 in their sera at different times, varying from 6 sequence (GGCGGG) for binding of the transcription to 15 or more months (Cole et al., 2003). It is reason- factor, Sp1. This site is also important for upregulation able to assume that PON1 levels are even lower in of PON1 by statins, and perhaps ethanol (Rao et al., fetuses, leading to concern about exposure of the 2003). Upregulation of PON1 expression by simvas- fetuses of mothers with low PON1 status to specific tatin involves an additional sequence in the 50 regula- compounds. tory region that represents a binding site for another Another related set of experiments examined the transcription factor, sterol regulatory element binding time course of appearance of the human PON1 allo- protein 2 (SREBP-2; Deakin et al., 2003). Deakin et al. forms in mice expressing hPON1Q192 and hPON1R192 (2003) also reported an interaction between Sp1 and on the PON1 null background (Cole et al., 2003). SREBP-2 in binding to the PON1 promoter. Interestingly, these studies showed that expression of Recently, the laboratory of Deborah Nickerson the human PON1 transgenes under control of the sequenced the entire PON1 genes from 47 individuals, human PON1 50 regulatory sequences followed the and identified more than 160 new polymorphisms, includ- same time course of appearance seen with mouse ing1novelcSNP,8newSNPsinthe50 regulatory region, PON1; i.e., the levels of the two hPON1 alloforms and12newSNPsinthe30 UTR (Jarvik et al., 2003). peaked around PND 21, whereas in humans, as noted Amongthenewgeneticvariations observedwasaW194X earlier, mature levels of expression are reached at 6–15 mutation which prompted us to reexamine DNA from months. These experiments indicated that the regula- plasma samples of individuals where there had been a tory elements controlling the developmental expres- discrepancy between their PON1Q192R genotyping results sion of PON1 are mostly conserved between mouse and their PON1 status as determined by the two-substrate and man. assay. One individual was identified among those samples There have been a number of reports related to the as having the W194X mutation (Fig. 1;Jarvik et al., 2003). influences of diet and environmental factors such as Re-sequencing the PON1 genes of other individuals who smoking on PON1 expression or activity. The reader is hadadiscrepancybetweentheirPON1Q192R genotypeand referred to a recent review of this topic for more PON1 status revealed three additional mutations, an information on this subject (Costa et al., submitted amino acid substitution at L90P, a probable deletion of for publication). part of the PON1Q192 allele and a mutation in an Asp codonthataffectedsplicing(Jarvik et al., 2003). Recent exploration of other polymorphisms that EPIDEMIOLOGICAL STUDIES AND PON1 affect the efficiency of RNA splicing (Faustino and LEVELS Cooper, 2003) suggest that the newly discovered poly- morphisms found in introns may well occur in splice To date, more than 100 studies have been carried out enhancer sequences and thus result in subtle, or even with the aim of determining whether there is a relation- significant, changes in PON1 expression levels. Deter- ship between genetic variations in PON1 and risk for mining which ones of these many intronic polymorph- specific diseases including vascular disease, vascular isms affect PON1 expression will be a difficult but disease among diabetics, Alzheimer’s disease and Par- most likely rewarding effort. kinson’s disease (reviewed in Brophy et al., 2002). The vast majority of these studies have considered only specific single nucleotide polymorphisms (SNPs) or DEVELOPMENTAL REGULATION OF PON1 haplotype associations, ignoring the importance of EXPRESSION variability in PON1 levels. This discussion is a plea for investigators to make use of the simple high- The observations described earlier provided convin- throughput determination of PON1 status via the cing evidence that PON1 plays a major role in deter- two-substrate analysis that was used to generate the C.E. Furlong et al. / NeuroToxicology 26 (2005) 651–659 657 data in Fig. 1. Many factors go into establishing an also clear from these studies that other pathways are individual’s PON1 status, but only some of these are important for detoxifying the parent compounds, as well currently known. Even if all known PON1 SNPs were as other OP compounds such as parathion and paraoxon. analyzed in an individual, it would not be possible to The availability of the PON1 knockout mice provides an predict the circulating levels of PON1 in their plasma, excellent system in which to investigate the contributions whereas the enzymatic determination of PON1 status of these other pathways without the interference of using the two-substrate assay provides the functional PON1.Forexample,with theseanimalsitwillbepossible PON1Q192R phenotype as well as the circulating levels to examine the contribution of the carboxylesterases, of PON1, the two factors involved in protecting an and cytochromes P450 in the detoxica- individual against toxic lipids and OP compounds. The tion of the oxon forms of the OPs. It is reasonable to same considerations bear on the questions of genetic predict that a ‘‘double knockout’’ animal missing both variability of PON1 and pharmacokinetics of drug PON1 and carboxylesterase will show increased sensi- metabolism through the PON1 pathway. tivity to chlorpyrifos oxon and diazoxon compared to the This discussion can also be viewed as a plea to the PON1knockoutanimalalone.Itwillalsobeinterestingto epidemiological community to avoid analyzing only determine whether increasing levels of specific cyto- PON1 SNP data. It will be more informative to analyze chromes P450 enhance or reduce toxicity. PON1 status for each individual in the study, and to The path to generating recombinant PON1 variants consider carefully not only the effects of the PON1Q192R for use in treating cases of OP poisoning has been made polymorphism, but also the relative contribution of other more clear with the recent elucidation of the three- metabolic pathways that might be involved in metaboliz- dimensional protein structure of a hybrid form of ing the toxin or drug of interest. The examples cited PON1 (Harel et al., 2004). This structure will facilitate earlier include cases where both PON1Q192R genotype the design of specific PON1 variants that can hydrolyze and PON1 levelsare important,suchaschlorpyrifosoxon nerve agents with sufficient catalytic efficiency to be of exposureorvasculardisease,andcaseswherePON1level use in treating cases of agent poisoning. In the future, it is the important factor governing sensitivity, such as might be possible to make use of recombinant PON1 to diazoxon exposure. The same will hold for many other reduce risk of vascular disease in individuals with low substrates of this promiscuous enzyme. PON1 status.

SUMMARY ACKNOWLEDGEMENTS

PON1 status plays an important role in protecting Research by the authors was supported by against exposures to diazinon and chlorpyrifos, parti- grants ES09601/EPA-R826886, ES09883, ES04696, cularly to the oxon residues present in these exposures. ES07033, ES11387, T32 AG0057, HL30568, HL67406. It is also clear that the developmental regulation of expression of PON1 is an important component of this picture. The most important conclusion to come from REFERENCES our studies is that to understand the role that PON1 plays in an individual’s sensitivity or resistance to a Adkins S, Gan KN, Mody M, La Du BN. Molecular basis for the given exposure or in the pharmacokinetic disposition polymorphic forms of human serum paraoxonase/arylesterase: of a specific drug, it is important to know both the glutamine or arginine at position 191, for the respective A or B levels of PON1 and in some cases the phenotype of the allozymes. Am J Hum Genet 1993;53:598–608. position 192 codon. These two important pieces of Aldridge WN. Serum I. Two types of esterase (A and B) information are readily extracted from a two-substrate hydrolysing p-nitrophenyl acetate, propionate and butyrate and a method for their determination. Biochem J 1953;53:110–7. determination of the individual’s plasma PON1 status. Aldridge WN. Serum esterases II. An enzyme hydrolysing diethyl p-nitrophenyl acetate (E600) and its identity with the A-esterase of mammalian sera. Biochem J 1953;53:117–24. FUTURE DIRECTIONS Aviram M, Hardak E, Vaya J, Mahmood S, Milo S, Hoffman A, et al. Human serum paraoxonase (PON1) Q and R selectively decrease lipid peroxides in human coronary and carotid arter- The studies performed to date have provided clear iosclerotic lesions. Circulation 2000;101:2510–7. evidence that PON1 plays a major role in detoxication of Biggadike K, Angell RM, Burgess CM, Farrell RM, Hancock AP, the oxon forms of diazinon and chlorpyrifos oxon. It is Harker AJ, et al. Selective plasma hydrolysis of glucocorti- 658 C.E. Furlong et al. / NeuroToxicology 26 (2005) 651–659

coids-gamma lactones and cyclic carbonates by the enzyme Draganov DI, La Du BN. Pharmacogenetics of : a paraoxonase: an ideal plasma inactivation mechanism. J Med brief review. Arch Pharmacol 2004;369:78–88. Chem 2000;43:19–21. Eckerson HW, White CM, La Du BN. The human serum para- Billecke S, Draganov D, Counsell R, Stetson P, Watson C, Hsu C, oxonase/arylesterase polymorphism. Am J Hum Genet et al. Human serum paraoxonase (PON1) Q and R 1983;35:1126–38. hydrolyze lactones and cyclic carbonate esters. Drug Metab Faustino NA, Cooper TA. Pre-mRNA splicing and human disease. Dispos 2000;28:1335–42. Genes Dev 2003;17:419–37. Blatter Garin MC, James RW, Dussoix P, Blanche M, Passa P, Geldmacher von Mallinckrodt M, Diepgen TL. The human serum Froguel P, et al. Paraoxonase polymorphism Met–Leu 54 is paraoxonase—polymorphism and specificity. Toxicol Environ associated with modified serum concentrations of the enzyme. Chem 1988;18:79–196. A possible link between the paraoxonase gene and increased Hassett C, Richter RJ, Humbert R, Chapline C, Crabb JW, Omie- risk of cardiovascular disease in diabetes. J Clin Invest cinski CJ, et al. Characterization of DNA clones encoding 1997;99:62–6. rabbit and human serum paraoxonase: the mature protein retains Brealey CB, Walker CH, Baldwin BC. A-Esterase activities in its signal sequence. Biochemistry 1991;30:10141–9. relation to the differential toxicity of pirimiphos-methyl to birds Harel M, Aharoni A, Gaidukov L, Brumshtein B, Khersonsky O, and mammals. Pest Sci 1980;11:546–54. Meged R, et al. Structure and evolution of the serum para- Brophy VH, Hastings MD, Clendenning JB, Richter RJ, Jarvik GP, oxonase family of detoxifying and anti-atherosclerotic Furlong CE. Polymorphisms in the human paraoxonase (PON1) . Nat Struct Mol Biol 2004;11:413–9. promoter. Pharmacogenetics 2001;11:77–84. Huff RA, Corcoran JJ, Anderson JK, Abou-Donia MB. Chlorpyr- Brophy VH, Jampsa RL, Clendenning JB, McKinstry LA, Furlong ifos oxon binds directly to muscarinic receptors and inhibits CE. Effects of 50 regulatory-region polymorphisms on paraox- cAMP accumulation in rat striatum. J Pharmacol Exp Ther onase-gene (PON1) expression. Am J Hum Genet 1994;269:329–35. 2001;68:1428–36. Humbert R, Adler DA, Disteche CM, Hassett C, Omiecinski CJ, Brophy VH, Jarvik GP, Furlong CE. PON1 polymorphisms. In: Furlong CE. The molecular basis of the human serum para- Costa LG, Furlong CE, editors. Paraoxonase (PON1) in health oxonase activity polymorphism. Nat Genet 1993;3:73–6. and disease: basic and clinical aspects. Norwell, MA: Kluwer Jarvik GP, Jampsa R, Richter RJ, Carlson CS, Rieder MJ, Nick- Academic Publishers; 2002. 53–77. erson DA, et al. Novel paraoxonase (PON1) nonsense and California Air Resources Board (CARB). Report for the application missense mutations predicted by functional genomic assay of and ambient air monitoring of chlorpyrifos (and the oxon PON1 status. Pharmacogenetics 2003;13:291–5. analog) in Tulare County during Spring/Summer, 1998. Project Kansouh ASH, Hopkins TL. Diazinon absorption, translocation, Nos. C96-040. Sacramento: Air Resources Board, California and metabolism in bean plants. J Agric Food Chem Environmental Protection Agency; 1998. 1968;16:446–50. Cole TB, Jampsa RL, Walter BJ, Arndt TL, Richter RJ, Shih DM, La Du BN, Furlong CE, Reiner E. Recommended nomenclature et al. Expression of human paraoxonase (PON1) during devel- system for the paraoxonases. Chem Biol Interact 1999;119– opment. Pharmacogenetics 2003;13:357–64. 120:599–601. Costa LG, Richter RJ, Murphy SD, Omenn GS, Motulsky AG, Leviev I, James RW. Promoter polymorphism of human paraox- Furlong CE. Species differences in serum paraoxonase correlate onase PON1 gene and serum paraoxonase activities and con- with sensitivity to paraoxon toxicity. In: Costa LG, Galli CL, centrations. Atheroscler Thromb Vasc Biol 2000;20:516–21. Murphy SD, editors. Toxicology of pesticides: experimental, Li WF, Costa LG, Furlong CE. Serum paraoxonase status: a major clinical and regulatory perspectives. Heidelberg: Springer-Ver- factor in determining resistance to organophosphates. J Toxicol lag; 1987. 262–6. Environ Health 1993;40:337–46. Costa LG, McDonald BE, Murphy SD, Richter RJ, Motulsky AG, Li WF, Furlong CE, Costa LG. Paraoxonase protects against Furlong CE. Serum paraoxonase and its influence on paraoxon chlorpyrifos toxicity in mice. Toxicol Lett 1995;76:219–26. and chlorpyrifos oxon toxicity in rats. Toxicol Appl Pharmacol Li WF, Costa LG, Richter RJ, Hagen T, Shih DM, Tward A, et al. 1990;103:66–76. Catalytic efficiency determines the in vivo efficacy of PON1 Costa LG, Vitalone A, Cole TB, Furlong CE. Modulation of for detoxifying organophosphates. Pharmacogenetics paraoxonase (PON1) activity. Biochem Pharmacol, submitted 2000;10:1–13. for publication. Mackness B, Mackness MI, Arrol T, Turkie W, Durrington PN. Costello J, Gibbons D, O’Connell L, Smith C, Bissell S, Carr J. Effect of the human serum paraoxonase 55 and 192 genetic Worker exposure to malathion, sulfur and nuisance dust during polymorphisms on the protection by high density lipoprotein date harvest. Sacramento: California Department of Food and against low density lipoprotein oxidative modifications. FEBS Agriculture, Division of Pest Management, Environmental Lett 1998;423:57–60. Protection and Worker Safety, Worker Health and Safety Main AR. The role of A-esterase in the acute toxicity of paraoxon, Branch; 1989. TEPP and parathion. Can J Biochem Physiol 1956;34:197–216. Davies H, Richter RJ, Kiefer M, Broomfield C, Sowalla J, Furlong Mazur A. An enzyme in animal tissue capable of hydrolyzing the CE. The human serum paraoxonase polymorphism is reversed phosphorus–fluorine bond of alkyl fluorophosphates. J Biol with diazoxon, soman and sarin. Nat Genet 1996;14:334–6. Chem 1946;164:271–89. Deakin S, Leviev I, Guernier S, James RW. Simvastatin modulates Mueller RF, Hornung S, Furlong CE, Anderson J, Giblett ER, expression of the PON1 gene and increases serum paraoxonase: Motulsky AG. Plasma paraoxonase polymorphism: a new a role for sterol regulatory element-binding protein-2. Arter- enzyme assay, population, family, biochemical and linkage ioscler Thromb Vasc Biol 2003;23:2083–9. studies. Am J Hum Genet 1983;35:393–408. C.E. Furlong et al. / NeuroToxicology 26 (2005) 651–659 659

Nigg HN, Allen JC, Brooks RF, Edwards GJ, Thompson NP, King Sorenson RC, Primo-Parmo SL, Kuo CL, Adkins S, Lockridge O, RW, et al. Dislodgeable residues of ethion in Florida citrus and La Du BN. Reconsideration of the catalytic center and mechan- relationships to weather variables. Arch Environ Contam Tox- ism of mammalian paraoxonase/arylesterase. Proc Natl Acad icol 1977;6:257–67. Sci USA 1995;92:7187–91. Ortigoza-Ferado J, Richter R, Hornung SK, Motulsky AG, Furlong Spear RC, Popendorf WJ, Leffingwell JT, Milby TH, Davies JE, CE. Paraoxon hydrolysis in human serum mediated by a Spencer WF. Fieldworker’s response to weathered residues of genetically variable arylesterase and albumin. Am J Hum Genet parathion. J Occup Med 1977;19:406–10. 1984;36:295–305. Spear RC, Popendorf WJ, Spencer WF, Milby TH. Worker Pond AL, Chambers HW, Chambers JE. detox- poisonings due to paraoxon residues. J Occup Med 1977;19: ication potential of various rat tissues via A-esterases and 411–4. aliesterase activities. Toxicol Lett 1995;78:245–52. Suehiro T, Nakamura T, Inoue M, Shiinoki T, Ikeda Y, Kumon Y, et Ralls JW, Gilmore DR, Cortes A. Fate of radioactive O,O-diethyl al. A polymorphism upstream from the human paraoxonase O-(2-iso-propyl-4-methylpyridmidin-6-yl) phosphorothioate (PON1) gene and its association with PON1 expression. Ather- on field-grown experimental crops. J Agric Food Chem osclerosis 2000;150:295–8. 1966;14:387–92. Tougou K, Nakamura A, Watanabe S, Okuyama Y, Morino A. Rao MN, Marmillot P, Gong M, Palmer DA, Seeff LB, Strader DB, Paraoxonase has a major role in the hydrolysis of prulifloxacin et al. Light, but not heavy alcohol drinking, stimulates para- (NM441), a prodrug of a new antibacterial agent. Drug Metab oxonase by upregulating liver mRNA in rats and humans. Dispos 1998;26:355–9. Metabolism 2003;52:1287–94. Wolfe HR, Armstrong JF, Staiff DC, Comer SW, Durham WF. Richter RJ, Furlong CE. Determination of paraoxonase (PON1) Exposure of apple thinners to parathion residues. Arch Environ status require more then genotyping. Pharmacogenetics Contam Toxicol 1975;3:257–67. 1999;9:745–53. Yuknavage KL, Fenske RA, Kalman DA, Keifer MC, Furlong CE. Shih DM, Gu L, Xia YR, Navab M, Li WF, Hama S, et al. Mice Simulated dermal contamination with capillary samples and lacking serum paraoxonase are susceptible to organophosphate field cholinesterase biomonitoring. J Toxicol Environ Health toxicity and . Nature 1998;394:284–7. 1997;51:35–55.